INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DO

-0.88

phabet

-0.77

 Dickinson

-0.73

beit

-0.72

MSG

-0.69

imester

-0.68

arget

-0.68

igon

-0.66

 Mellon

-0.66

CLASSIFIED

-0.66

POSITIVE LOGITS

Cth

0.80

roots

0.68

unts

0.66

trop

0.65

forms

0.64

soc

0.62

gets

0.61

 perman

0.60

 basics

0.60

urg

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.